On Wed, Jan 06, 2010 at 12:19:56PM -0500, Hudes, Dana wrote: > >the ISP feature of VM would allow you to drill down to individual spindles > >and place subdisks on each spindle. > > Individual spindles of the RAID group? Doesn't that defeat the purpose of the > RAID group? > Striping across LUNs gets ...interesting; we usually just use them concat. Of > course that's with a real SAN array such as Hitachi 99x0 or Sun 61x0. > I'm not sure I see the point of striping LUNs. If you are having performance > problems from the array, fix the layout of the RAID group on the array: > that's why you pay the big bucks to Hitachi for their hardware. I not sure I > want to know about the load that could flatline a RAID-6 array of 6 15K RPM > Fiber channel disks backed by a multigigabyte RAM cache.
Dana, I assure you that we did a lot of thorough performance tests of each raid group to have real numbers so our decision to stripe is made after careful study. We spent a lot of time on it. E.g. we got the following results: (7+1)*4 ~ 6242 IOPS (4+4) ~ 2294 IOPS At first sight (7+1)*4 is faster. But when you compare number of disks involved in each raid you get: (7+1)*4 ~ 6242 IOPS = 195 IOPS/HDD (4+4) ~ 2294 IOPS = 287 IOPS/HDD So if we stripe over (4+4)*3 we can get better result then using (7+1)*4 but using less disks (!) And array cache (which is _not_ infinite ...) helps but up to some point. When you have stream of constant data, the cache gets saturated and destaging starts. So after all you get back to basic entity which is hard disk. This is why we decided to use striping. But still would like to preserve ability to increase number of columns. Just in case. That is why I send the first e-mail Regards Przemyslaw Bak (przemol) -- http://przemol.blogspot.com/ ---------------------------------------------------------------------- Sypie, sypie, ale czy sypac bedzie nadal? Sprawdz pogode >> http://link.interia.pl/f2530 _______________________________________________ Veritas-vx maillist - Veritas-vx@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-vx
